A simulation study using the hybrid statistic and the meta-analysis t-test for data with matched and unmatched subjects in the interim analysis of clinical trials

An interim analysis is usually applied in later phase II or phase III trials to find convincing evidence of a significant treatment difference that may lead to trial termination at an earlier point than planned at the beginning. This can result in the saving of patient resources and shortening of drug development and approval time. In addition, ethics and economics are also the reasons to stop a trial earlier. In clinical trials of eyes, ears, knees, arms, kidneys, lungs, and other clustered treatments, data may include distribution-free random variables with matched and unmatched subjects in one study. It is important to properly include both subjects in the interim and the final analyses so that the maximum efficiency of statistical and clinical inferences can be obtained at different stages of the trials. So far, no publication has applied a statistical method for distribution-free data with matched and unmatched subjects in the interim analysis of clinical trials. In this simulation study, the hybrid statistic was used to estimate the empirical powers and the empirical type I errors among the simulated datasets with different sample sizes, different effect sizes, different correlation coefficients for matched pairs, and different data distributions, respectively, in the interim and final analysis with 4 different group sequential methods. Empirical powers and empirical type I errors were also compared to those estimated by using the meta-analysis t-test among the same simulated datasets. Results from this simulation study show that, compared to the meta-analysis t-test commonly used for data with normally distributed observations, the hybrid statistic has a greater power for data observed from normally, log-normally, and multinomially distributed random variables with matched and unmatched subjects and with outliers. Powers rose with the increase in sample size, effect size, and correlation coefficient for the matched pairs. In addition, lower type I errors were observed estimated by using the hybrid statistic, which indicates that this test is also conservative for data with outliers in the interim analysis of clinical trials.^

Ethnic Minority Children’s Active Commuting to School and Association with Physical Activity and Pedestrian Safety Behaviors

Background: Children's active commuting to school, i.e. walking or cycling to school, was associated with greater moderate-to-vigorous physical activity, although studies among ethnic minorities are sparse. Objectives: Among a low-income, ethnic minority sample of fourth grade students from eight public schools, we examined (1) correlates of active commuting to school and (2) the relationship between active commuting to school and moderate-to-vigorous physical activity. Methods: We conducted a cross-sectional analysis of baseline measurements from a sample of participants (n=149) aged 9-12 years from a walk to school intervention study in Houston, Texas. The primary outcome was the weekly rate of active commuting to school. Daily moderate-to-vigorous physical activity, measured by accelerometers, was a secondary outcome. Child self-efficacy (alpha=0.75), parent self-efficacy (alpha=0.88), and parent outcome expectations (alpha=0.78) were independent variables. Participant characteristics (age, gender, race/ethnicity, distance from home to school, acculturation, and BMI percentile) were independent sociodemographic variables. We used mixed-model regression analyses to account for clustering by school and a stepwise procedure with backward elimination of non-significant interactions and covariates to identify significant moderators and predictors. School-level observations of student pedestrians were assessed and compared using chi-square tests of independence. Results: Among our sample, which was 61.7% Latino, the overall rate of active commuting to school was 43%. In the mixed model for active commuting to school, parent self-efficacy (std. beta = 0.18, p=0.018) and age (std. beta = 0.18, p=0.018) were positively related. Latino students had lower rates of active commuting to school than non-Latinos ( 16.5%, p=0.040). Distance from home to school was inversely related to active commuting to school (std. beta = 0.29, p<0.001). In the mixed model for moderate-to-vigorous physical activity, active commuting to school was positively associated (std. beta = 0.31, p <0.001). Among the Latino subsample, child acculturation was negatively associated with active commuting to school (std. beta = -0.23, p=0.01). With regard to school-level pedestrian safety observations, 37% of students stopped at the curb and 2.6% looked left-right-left before crossing the street. Conclusion: Although still below national goals, the rate of active commuting was relatively high, while the rate of some pedestrian safety behaviors was low among this low-income, ethnic minority population. Programs and policies to encourage safe active commuting to school are warranted and should consider the influence of parents, acculturation, and ethnicity.

Changes in serum lipids with change and fluctuation in body mass index and body fat distribution in Mexican Americans with type 2 diabetes

This dissertation was written in the format of three journal articles. Paper 1 examined the influence of change and fluctuation in body mass index (BMI) over an eleven-year period, on changes in serum lipid levels (total, HDL, and LDL cholesterol, triglyceride) in a population of Mexican Americans with type 2 diabetes. Linear regression models containing initial lipid value, BMI and age, BMI change (slope of BMI), and BMI fluctuation (root mean square error) were used to investigate associations of these variables with change in lipids over time. Increasing BMI over time was associated with gains in total and LDL cholesterol and triglyceride levels in women. Fluctuation of BMI was not associated with detrimental lipid profiles. These effects were independent of age and were not statistically significant in men. In Mexican-American women with type 2 diabetes, weight reduction is likely to result in more favorable levels of total and LDL cholesterol and triglyceride, without concern for possible detrimental effects of weight fluctuation. Weight reduction may not be as effective in men, but does not appear to be harmful either. ^ Paper 2 examined the associations of upper and total body fat with total cholesterol, HDL and LDL cholesterol, and triglyceride levels in the same population. Multilevel analysis was used to predict serum lipid levels from total body fat (BMI and triceps skinfold) and upper body fat (subscapular skinfold), while controlling for the effects of sex, age and self-correlations across time. Body fat was not strikingly associated with trends in serum lipid levels. However, upper body fat was strongly associated with triglyceride levels. This suggests that loss of upper body fat may be more important than weight loss in management of the hypertriglyceridemia commonly seen in type 2 diabetes. ^ Paper 3 was a review of the literature reporting associations between weight fluctuation and lipid levels. Few studies have reported associations between weight fluctuation and total, LDL, and HDL cholesterol and triglyceride levels. The body of evidence to date suggests that weight fluctuation does not strongly influence levels of total, LDL and HDL cholesterol and triglyceride. ^

Bayesian generalized linear models for meta-analysis of diagnostic tests

With the recognition of the importance of evidence-based medicine, there is an emerging need for methods to systematically synthesize available data. Specifically, methods to provide accurate estimates of test characteristics for diagnostic tests are needed to help physicians make better clinical decisions. To provide more flexible approaches for meta-analysis of diagnostic tests, we developed three Bayesian generalized linear models. Two of these models, a bivariate normal and a binomial model, analyzed pairs of sensitivity and specificity values while incorporating the correlation between these two outcome variables. Noninformative independent uniform priors were used for the variance of sensitivity, specificity and correlation. We also applied an inverse Wishart prior to check the sensitivity of the results. The third model was a multinomial model where the test results were modeled as multinomial random variables. All three models can include specific imaging techniques as covariates in order to compare performance. Vague normal priors were assigned to the coefficients of the covariates. The computations were carried out using the 'Bayesian inference using Gibbs sampling' implementation of Markov chain Monte Carlo techniques. We investigated the properties of the three proposed models through extensive simulation studies. We also applied these models to a previously published meta-analysis dataset on cervical cancer as well as to an unpublished melanoma dataset. In general, our findings show that the point estimates of sensitivity and specificity were consistent among Bayesian and frequentist bivariate normal and binomial models. However, in the simulation studies, the estimates of the correlation coefficient from Bayesian bivariate models are not as good as those obtained from frequentist estimation regardless of which prior distribution was used for the covariance matrix. The Bayesian multinomial model consistently underestimated the sensitivity and specificity regardless of the sample size and correlation coefficient. In conclusion, the Bayesian bivariate binomial model provides the most flexible framework for future applications because of its following strengths: (1) it facilitates direct comparison between different tests; (2) it captures the variability in both sensitivity and specificity simultaneously as well as the intercorrelation between the two; and (3) it can be directly applied to sparse data without ad hoc correction. ^

Relationship between dietary patterns and body mass index in Hispanic children ages 4--19 years of age

Background. In over 30 years, the prevalence of overweight for children and adolescents has increased across the United States (Barlow et al., 2007; Ogden, Flegal, Carroll, & Johnson, 2002). Childhood obesity is linked with adverse physiological and psychological issues in youth and affects ethnic/minority populations in disproportionate rates (Barlow et al., 2007; Butte et al., 2006; Butte, Cai, Cole, Wilson, Fisher, Zakeri, Ellis, & Comuzzie, 2007). More importantly, overweight in children and youth tends to track into adulthood (McNaughton, Ball, Mishra, & Crawford, 2008; Ogden et al., 2002). Childhood obesity affects body functions such as the cardiovascular, respiratory, gastrointestinal, and endocrine systems, including emotional health (Barlow et al., 2007, Ogden et al., 2002). Several dietary factors have been associated with the development of obesity in children; however, these factors have not been fully elucidated, especially in ethnic/minority children. In particular, few studies have been done to determine the effects of different meal patterns on the development of obesity in children. Purpose. The purpose of the study is to examine the relationships between daily proportions of energy consumed and energy derived from fat across breakfast, lunch, dinner, and snack, and obesity among Hispanic children and adolescents. Methods. A cross-sectional design was used to evaluate the relationship between dietary patterns and overweight status in Hispanic children and adolescents 4-19 years of age who participated in the Viva La Familia Study. The goal of the Viva La Familia Study was to evaluate genetic and environmental factors affecting childhood obesity and its co-morbidities in the Hispanic population (Butte et al., 2006, 2007). The study enrolled 1030 Hispanic children and adolescents from 319 families and examined factors related to increased body weight by focusing on a multilevel analysis of extensive sociodemographic, genetic, metabolic, and behavioral data. Baseline dietary intakes of the children were collected using 24-hour recalls, and body mass index was calculated from measured height and weight, and classified using the CDC standards. Dietary data were analyzed using a GEE population-averaged panel-data model with a cluster variable family identifier to include possible correlations within related data sets. A linear regression model was used to analyze associations of dietary patterns using possible covariates, and to examine the percentage of daily energy coming from breakfast, lunch, dinner, and snack while adjusting for age, sex, and BMI z-score. Random-effects logistic regression models were used to determine the relationship of the dietary variables with obesity status and to understand if the percent energy intake (%EI) derived from fat from all meals (breakfast, lunch, dinner, and snacks) affected obesity. Results. Older children (age 4-19 years) consumed a higher percent of energy at lunch and dinner and less percent energy from snacks compared to younger children. Age was significantly associated with percentage of total energy intake (%TEI) for lunch, as well as dinner, while no association was found by gender. Percent of energy consumed from dinner significantly differed by obesity status, with obese children consuming more energy at dinner (p = 0.03), but no associations were found between percent energy from fat and obesity across all meals. Conclusions. Information from this study can be used to develop interventions that target dietary intake patterns in obesity prevention programs for Hispanic children and adolescents. In particular, intervention programs for children should target dietary patterns with energy intake that is spread throughout the day and earlier in the day. These results indicate that a longitudinal study should be used to further explore the relationship of dietary patterns and BMI in this and other populations (Dubois et al., 2008; Rodriquez & Moreno, 2006; Thompson et al., 2005; Wilson et al., in review, 2008). ^

The association between mental health and hypertension in the 2005 Department of Defense Population Survey

Major objectives within Healthy People 2010 include improving hypertension and mental health management of the American population. Both mental health issues and hypertension exist in the military which may decrease the health status of military personnel and diminish the ability to complete assigned missions. Some cases may be incompatible with military service even with optimum treatment. In the interest of maintaining a fit fighting force, the Department of Defense regularly conducts a survey of health related behaviors among active duty military personnel. The 2005 DoD Survey was conducted to obtain information regarding health and behavioral readiness among active duty military personnel to assess progress toward selected Healthy People 2010 objectives. ^ This study is a cross-sectional prevalence design looking at the association of hypertension treatment with mental health issues (either treatment or perceived need for treatment) within the military population sampled in the 2005 DoD Survey. There were 16,946 military personnel in the final cross-sectional sample representing 1.3 million active duty service members. The question is whether there is a significant association between the self-reported occurrence of hypertension and the self-reported occurrence of mental health issues in the 2005 DoD Survey. In addition to these variables, this survey examined the contribution of various sociodemographic, occupational, and behavioral covariates. An analysis of the demographic composition of the study variables was followed by logistic analysis, comparing outcome variables with each of the independent variables. Following univariate regression analysis, multivariate regression was performed with adjustment (for those variables with an unadjusted alpha level less than or equal to 0.25). ^ All the mental health related indicators were associated with hypertension treatment. The same relationship was maintained after multivariate adjustment. The covariates remaining as significant (p < 0.05) in the final model included gender, age, race/ethnicity and obesity. There is a need to recognize and treat co-morbid medical diagnoses among mental health patients and to improve quality of life outcomes, whether in the military population or the general population. Optimum health of the individual can be facilitated through discovery of treatable cases, to minimize disruptions of military missions, and even allow for continued military service. ^

Variation in hospital inpatient charges by payer in Houston, Texas

An analysis of variation in hospital inpatient charges in the greater Houston area is conducted to determine if there are consistent differences among payers. Differences in charges are examined for 59 Composite Diagnosis Related Groups (CDRGs) and two regression equations estimating charges are specified. Simple comparison of mean charges by diagnostic categories are significantly different for 42 (71 percent) of the 59 categories examined. In 41 of the 42 significant categories, charges to Medicaid were less than charges to private insurers. Meta-analytic statistical techniques yielded a weighted average effect size of $-$0.7198 for the 59 diagnostic categories, indicating an overall effect that Medicaid charges were less than private insurance charges. Results of a multiple regression estimating charges showed that private insurance was a significant independent variable, along with age, length of stay, and hospital variables. Results indicated consistent differential charges in the present analysis. ^

Data quality in trauma transfusion studies and the impact of missing data on predicting massive transfusion

Maximizing data quality may be especially difficult in trauma-related clinical research. Strategies are needed to improve data quality and assess the impact of data quality on clinical predictive models. This study had two objectives. The first was to compare missing data between two multi-center trauma transfusion studies: a retrospective study (RS) using medical chart data with minimal data quality review and the PRospective Observational Multi-center Major Trauma Transfusion (PROMMTT) study with standardized quality assurance. The second objective was to assess the impact of missing data on clinical prediction algorithms by evaluating blood transfusion prediction models using PROMMTT data. RS (2005-06) and PROMMTT (2009-10) investigated trauma patients receiving ≥ 1 unit of red blood cells (RBC) from ten Level I trauma centers. Missing data were compared for 33 variables collected in both studies using mixed effects logistic regression (including random intercepts for study site). Massive transfusion (MT) patients received ≥ 10 RBC units within 24h of admission. Correct classification percentages for three MT prediction models were evaluated using complete case analysis and multiple imputation based on the multivariate normal distribution. A sensitivity analysis for missing data was conducted to estimate the upper and lower bounds of correct classification using assumptions about missing data under best and worst case scenarios. Most variables (17/33=52%) had <1% missing data in RS and PROMMTT. Of the remaining variables, 50% demonstrated less missingness in PROMMTT, 25% had less missingness in RS, and 25% were similar between studies. Missing percentages for MT prediction variables in PROMMTT ranged from 2.2% (heart rate) to 45% (respiratory rate). For variables missing >1%, study site was associated with missingness (all p≤0.021). Survival time predicted missingness for 50% of RS and 60% of PROMMTT variables. MT models complete case proportions ranged from 41% to 88%. Complete case analysis and multiple imputation demonstrated similar correct classification results. Sensitivity analysis upper-lower bound ranges for the three MT models were 59-63%, 36-46%, and 46-58%. Prospective collection of ten-fold more variables with data quality assurance reduced overall missing data. Study site and patient survival were associated with missingness, suggesting that data were not missing completely at random, and complete case analysis may lead to biased results. Evaluating clinical prediction model accuracy may be misleading in the presence of missing data, especially with many predictor variables. The proposed sensitivity analysis estimating correct classification under upper (best case scenario)/lower (worst case scenario) bounds may be more informative than multiple imputation, which provided results similar to complete case analysis.^

Rigorously assessing whether the data backs the back school.

A rigorous between-subjects methodology employing independent random samples and having broad clinical applicability was designed and implemented to evaluate the effectiveness of back safety and patient transfer training interventions for both hospital nurses and nursing assistants. Effects upon self-efficacy, cognitive, and affective measures are assessed for each of three back safety procedures. The design solves the problem of obtaining randomly assigned independent controls where all experimental subjects must participate in the training interventions.

Commentary on "Child Welfare Waivers: The Stakes for Families"

The purpose of this piece is to provide commentary of an article, Child Welfare Waivers: The Stakes for Your State, that discusses the recent reauthorization of the Title IV-E Child Welfare Waivers. The article provides an overview of funds available to the states for child welfare programs and their intended purpose and restrictions placed on use. As structured, the present system rewards states monetarily for maintaining foster care. Research from waiver programs shows promising results for improved outcomes at the same or lower financial cost by utilizing safe, proven alternatives to the current foster care system. Waiver funds also protect the financial commitment to child welfare because state legislative budget slashing in this area will result in the loss of Federal funding. The independent analysis required with the grant of a waiver must be maintained to provide ongoing analysis and oversight of the increase spending flexibility. Stakeholders must be aware of the program and its results and use these funds as an opportunity to assess new concepts and apply programs best suited to the needs of children in their state. Allowing those “on the ground” to determine appropriate programming and careful result assessment may be the best means for protecting children, preserving families and doing both in a manner that makes the most efficient use of available resources.

Inputs to parasol ganglion cells in the primate retina

Retinal ganglion cells carry signals from the eye to the brain. One of the most common types of ganglion cells is parasol cells. They have larger dendritic trees, somas and axons than other ganglion cells. While much was known about parasol cell light responses, little was known about how these responses are formed. One possibility is that they receive input from a unique set of local circuit neurons that have similar responses. The goal was to identify these presynaptic neurons and study their synaptic connectivity.^ Ganglion cells receive input from bipolar and amacrine cells, but there are numerous subtypes of each. To determine which of these were most likely to provide input to parasol cells, the parasol cells were intracellularly-injected and then various bipolar and amacrine cells were immunolabeled and the tissue analyzed using a confocal microscope. DB3 bipolar cells labeled with antibodies to calbindin made extensive contacts with OFF parasol cells. Antibodies to recover in labeled flat midget bipolar cells (FMB). They made only random contacts with OFF parasol cells, and they are not expected to provide significant input. Type DB2 bipolar cells and FMB cells labeled with antibodies to excitatory amino acid transporter-2 made extensive contacts with OFF parasol cells. This suggests that DB2 bipolar cells are likely to provide input to parasol cells.^ Two types of amacrine cells were labeled in material containing injected parasol cells. Cholinergic amacrine cells were labeled with antibodies to choline acetyltransferase, and they made extensive contacts with ON parasol cells. The large amacrine cells labeled with antibodies to a precursor of cholecystokinin were among the amacrine cells that are tracer-coupled to parasol cells.^ From electron microscopic (EM) analysis, most of the synapses made by DB3 axons were found on varicosities. Some postsynaptic and presynaptic amacrine cells resembled AII amacrine cells. Others were relatively electron-lucent and may be cholinergic amacrine cells or cholecystokinin-containing amacrine cells. Gap junctions were found between neighboring DB3 axons. They occurred whenever two axons contacted each other, and the junctions were as large as the area of contact. In double-label EM experiments, DB3 axons made synapses onto OFF parasol cells. ^

GENERAL WELL-BEING AND USE OF SOCIAL SUPPORT

Research interest on well-being and social support has focused largely on social factors as related to attaining and maintaining well-being, self-perceptions of well-being and to a lesser extent the relationship of current level of self-perceived well-being to use of formal or informal sources of social support. This study analyzed responses to the General Well-Being Schedule of 6,913 subjects (25-74 years) interviewed during the National Health and Nutrition Examination Survey (1971-1975). The purpose of this analysis was to relate the level of GWBS scores to the use of social support, both informal (family and friends) and formal (community professionals).^ Study questions addressed were whether well-being level was related to selection of a specific social support resource and/or rate of use of resources and whether gender differences were apparent in level of well-being and social support use. Because age, sex, race, socioeconomic status (income and education) and marital status may confound the relation between level of GWB and type of social support chosen, the association between these variables with GWB and use of social support were considered. For analysis, test scores were grouped into four categories and for detailed analysis, two categories: low (0-70) and high (71-110). Cross tabulations and percentages were computed and the chi-square test of significance was used.^ Although 16 to 25 percent of the sample population reported low well-being, less than 10 percent used formal resources to discuss emotional, mental or behavior problems. Medical resources, mostly physicians, were the most used formal social supports. Informal social support was important for all well-being levels where 65-77% of each category reported using this resource.^ While well-being level does not appear to serve as a screener/selector of type of formal social support used, it is related to rates of use. Females reported slightly lower well-being than males, and except in the lowest well-being group, had higher rates of social support use. Findings support the conclusion that perceived well-being is related to use of social support such that the lower the well-being, the greater tendency to use formal and/or informal social support. ^

Predictors of HIV testing among injection drug users

The current analysis examined the association of several demographic and behavioral variables with prior HIV testing within a population of injection drug users (IDUs) living in Harris County, Texas in 2005 (n=563). After completing the initial univariate analyses of all potential predictors, a multivariable model was created. This model was designed to guide future intervention efforts. Data used in this analysis were collected by the University of Texas School of Public Health in association with the Houston Department of Health and Human Services for the first IDU cycle of the National HIV Behavioral Surveillance System. About 76% of the IDUs reported previously being tested for HIV. Demographic variables that displayed a significant association with prior testing during the univariate analyses include age, race/ethnicity, birth outside the United States, education level, recent arrest, and current health insurance coverage. Several drug-related and sexual behaviors also demonstrated significant associations with prior testing, including age of first injection drug use, heroin use, methamphetamine use, source of needles or syringes, consistent use of new needles, recent visits to a shooting gallery or similar location, previous alcohol or drug treatment, condom use during their most recent sexual encounter, and having sexual partners who also used injection drugs. Additionally, the univariate analyses revealed that recent use of health or HIV prevention services was associated with previously testing for HIV. The final multivariable model included age, race/ethnicity, recent arrest, previous alcohol or drug treatment, and heroin use. ^

Application of cell line based genomic predictors to predict response to targeted therapies in breast cancer

Cancer cell lines can be treated with a drug and the molecular comparison of responders and non-responders may yield potential predictors that could be tested in the clinic. It is a bioinformatics challenge to apply the cell line-derived multivariable response predictors to patients who respond to therapy. Using the gene expression data from 23 breast cancer cell lines, I developed three predictors of dasatinib sensitivity by selecting differentially expressed genes and applying different classification algorithms. The performance of these predictors on independent cell lines with known dasatinib response was tested. The predictor based on weighted voting method has the best overall performance. It correctly predicted dasatinib sensitivity in 11 out of 12 (92%) breast and 17 out of 23 (74%) lung cancer cell lines. These predictors were then applied to the gene expression data from 133 breast cancer patients in an attempt to predict how the patients might respond to dasatinib therapy. Two predictors identified 13 patients in common to be dasatinib sensitive. Sixty two percent of these cases are triple negative (ER-negative, HER2-negative and PR-negative) and 76% are double negative. The result is consistent with the findings from other studies, which identified a target population for dasatinib treatment to be triple negative or basal breast cancer subtype. In conclusion, we think that the cell line-derived dasatinib classifiers can be applied to the human patients. ^

Environmental influences on self-reported leisure-time physical activity in the Austin/Travis County STEPS survey 2005--2009

This study was an examination of environmental and social correlates of physical inactivity in a socio-economically disadvantaged community. This study was a secondary analysis of data collected by the Austin/Travis County Department of Health and Human Services. The study included an analysis of perceived environmental influences (e.g. access to physical support features like sidewalks and streetlights). This study also investigated several features of the social environment (e.g. perceived neighborhood crime and social influence). Participants’ beliefs and attitudes about the neighborhood were investigated. Results included estimates of the association between neighborhood factors and physical inactivity controlling for age, gender and education. This study found significant associations for social and environmental variables with physical inactivity. The goal of this work was to identify factors that contribute to inactivity and address a number of environmental and neighborhood risk factors that contribute to sedentary behaviors in a population of relative social and economic disadvantage.^